The Design And Implement Of A Mobiles' Chinese Input System Based On Statistical Language Model

Intelligent Chinese input technology (ICIT) has been widely used in desktop computers, by researching the characteristics of the Chinese input on mobiles, this paper imports the ICIT to the mobile platform, and implements the sentential forecast by calculating the duality probability.First,this thesis describes a improved method which digital input method, introduce the n-gram model which is widely used in intelligent input technology and two common smoothing method, and presents a self-study model which can renmember user's input habit. Then, the paper fully describes the storage structure and correlative search arithmetic of the data files used by the Chinese input system. .When designing these data structure and arithmetic, the limitation of the moblies was fully considered. Base on the analysis above, the paper emphasis on the forecast arithmetic of sentential input, which uses the positive-integer split table to improve the performance. After introduces the input method and the characteristics of the Symbian platform, the paper describes the progress of implement the input system above, and refers to the tests performance of it.According to the characteristics of, the text input on the mobiles, this paper uses the movie dialog as the statistic source of the data file. The experimental results show that in the oral condition, the system gets a preferable performance. This paper applies the statistical language model to the mobile's input system, which will be of great practical significance.